Best way to read big data from SQL Server with pandas

Question

I\'m looking to read millions of rows from a ODBC connection (SQL Server), and perform some processing with Pandas. If I try to directly load the dataframe in memory, I enco